Runoff and Erosion in a Piñon–Juniper Woodland: Influence of Vegetation Patches

which generally have little or no vegetation, produce runoff; sink areas, located downslope of the source arIn many semiarid regions, runoff and erosion differ according to eas, receive and store the runoff and thereby become vegetation patch type. These differences, although hypothesized to fundamentally affect ecological processes, have been poorly quantienriched and relatively productive. The results of sevfied. In a semiarid piñon–juniper woodland [Pinus edulis Engelm. eral theoretical studies suggest that the transfer of water and Juniperus monosperma (Engelm.) Sarg.] in northern New Mexico, and nutrients through this process is important both we measured runoff and erosion from the three patch types that ecologically and hydrologically (Mauchamp et al., 1994; compose these woodlands: Canopy patches (those beneath woody Ludwig and Marsden, 1995; Reynolds et al., 1997; Davplants), vegetated patches in intercanopy areas, and bare patches in enport et al., 1998; Aguiar and Sala, 1999; Dunkerley, intercanopy areas. The bare intercanopy patches exhibited the highest 1999; Klausmeier, 1999). Although widely recognized rates, followed by vegetated intercanopy patches and then by canopy as important, the transfer and redistribution of water patches. Large convective summer storms, though relatively infreand sediment among vegetation patch types in semiarid quent, generated much of the runoff and most of the sediment; prolonged frontal storms were capable of generating considerable runoff landscapes have rarely been directly measured. Most of but little sediment. A portion of the runoff and most of the sediment the information we have to date is based on indirect generated from bare intercanopy patches was redistributed downindicators of runoff or other ecologically relevant indislope, probably to adjacent vegetated intercanopy patches, demonces, such as changes in soil water content (Cornet et al., strating connectivity between these two patch types. Our results indi1992; Bromley et al., 1997) and in soil nutrients (Ludwig cate that there are significant and important differences in runoff and and Tongway, 1995), distribution of tree seedlings sediment production from the three patch types; that bare intercanopy (Montaña et al., 1990), degree of plant water stress patches act as sources of both water and sediment for the vegetated (Schlesinger et al., 1989; Anderson and Hodgkinson, intercanopy patches; and that the transfer of water and sediment at 1997), and geomorphic characteristics (Greene, 1992; small scales is both frequent enough and substantial enough to be considered ecologically significant. Hysell and Grier, 1996; Wondzell et al., 1996). Patchscale measurements of runoff, but not of runon, in grasslands and desert scrub are presented by Schlesinger et al. (1999); the data are primarily from rainfall simulation S landscapes can be viewed as a mosaic of but measurements from actual precipitation are also vegetation patches at practically every scale of obreported. Direct measurements of both runoff and servation. The different patterns formed by these runon from actual precipitation would provide a valupatches generally reflect differences in soil water available complement to these studies. They would also enability, which are traceable to myriad factors—such as able runoff processes at the patch scale to be related to aspect, slope, parent material, soil properties, microclithose at the hillslope scale. matic effects, and surface runoff characteristics. Landscape ecologists are evaluating the health or The interrelationships between vegetation patterns functionality of semiarid rangelands on the basis of inand the distribution of surface runoff have been recogteractions between runoff and vegetation (Ludwig and nized in semiarid regions throughout the world, includTongway, 1995; Ludwig et al., 1997). A fully functional ing Australia (Smith and Morton, 1990; Dunkerley and Brown, 1995; Ludwig et al., 1999), Mexico (Montaña et semiarid ecosystem is defined as one in which only a al., 1990; Cornet et al., 1992), Niger (White, 1971; Bromvery small part of the water and nutrients that enter the ley et al., 1997), and the USA (Schlesinger et al., 1989; system are subsequently lost. Runoff, when it occurs, is Seyfried, 1991; Wondzell et al., 1996; Schlesinger et al., redistributed within the system and effectively trapped 1999). These interrelationships have profound ecologiand stored locally. A dysfunctional ecosystem is one cal implications, as outlined by Ludwig et al., 1997. from which a significant portion of the water and nutriRedistribution of water and other resources by surents are being lost, generally because the network of face runoff may be especially important at small scales vegetation patches is too spotty to trap surface runoff. (,10 m2). In semiarid regions it is now recognized that Practices such as overgrazing can change a functional at small scales, vegetation patches may function either ecosystem into a dysfunctional one: As the number and as sources or sinks (Ludwig et al., 1997). Source areas, size of vegetated patches are reduced, more and more water and resources are carried away from the system Kevin D. Reid, Bradford P. Wilcox, and David D. Breshears, Environ(Schlesinger et al., 1990; Ludwig et al., 1997, 1999; mental Science Group, MS J495, Los Alamos National Lab., Los Schlesinger et al., 1999). Alamos, NM 87545; Kevin D. Reid and Lee MacDonald, Dep. of In arid and semiarid shrublands and woodlands, a Earth Resources, Colorado State Univ., Fort Collins, CO 80523; and fundamental dichotomy exists between the areas beBradford P. Wilcox, Inter-American Institute for Global Change Research, Av. Astronautas 1758, 12227-010, Sao Jose dos Campos, SP neath the canopies of woody plants and the intercanopy Brazil. Received 20 July 1998. *Corresponding author (daveb@ areas that separate them (Breshears and Barnes, 1999). lanl.gov). The characteristics that differentiate canopy patches from intercanopy patches are well-documented: (i) the Published in Soil Sci. Soc. Am. J. 63:1869–1879 (1999).